Skip to content

Releases: nextbridgehq/project-charter

Release list

v0.1.0 — Initial Public Release

v0.1.0 — Initial Public Release Latest
Latest

Choose a tag to compare

View all tags
@nextbridgehq nextbridgehq released this 27 Aug 14:17
v0.1.0
b61a370

🚀 Project Charter v0.1.0 — Initial Public Release

We are excited to announce the first public release of Project Charter, an open-source AI Agent Governance Engine built by Nextbridge.

What is Project Charter?

Modern AI coding agents are powerful — but unpredictable. They can rewrite architecture boundaries, ignore CI policies, or silently skip human review steps. Project Charter solves this by turning your project's conventions into a strictly-enforced, stateful runtime that agents must operate within.

Instead of giving an AI a massive wall of text and hoping it behaves, Project Charter models complex coding workflows as a deterministic State Machine governed by Machine-Readable Contracts.

✨ What's Included in v0.1.0

🔒 Runtime Convention Enforcement

Project Charter parses your CONVENTIONS.md or AGENTS.md file and enforces the rules at runtime — before any file is touched. Architectural boundaries, import restrictions, and git policies are all enforceable.

charter enforce
charter enforce --format=ci --strict  # for GitHub Actions inline annotations

🧠 Stateful Workflow Engine

Workflows run as a strict state machine with phases (PENDING → IN_PROGRESS → AWAITING_APPROVAL → COMPLETED). State is automatically persisted into .charter/<run_id>.json inside your repo, making every workflow fully resumable after a crash or pause.

🛑 Human Approval Gates

Critical workflow phases can be gated behind a mandatory human review step. The engine pauses the AI until a human explicitly approves:

charter approve

🔌 MCP Server — Native Agent Integration

Project Charter ships with a built-in Model Context Protocol (MCP) server, making it natively compatible with Claude Code, Cursor, Gemini, and Codex out of the box — no middleware required.

uvx --from "project-charter[mcp]" charter-mcp

🤖 Multi-Provider Adapters

Swappable provider adapters for:

  • Anthropic (Claude)
  • OpenAI (Codex / GPT)
  • Google (Gemini)

🔎 Capability Auditing

Every tool call an agent makes is logged. Inspect exactly what an AI agent did, what it tried to access, and what was blocked:

charter audit

📦 Bundled Governance Skills

Four ready-to-use AI skills are bundled with the package at v0.1.0:

  • charter-analysis — Deep codebase analysis and architecture mapping
  • charter-governance — Extract and enforce coding standards from existing code
  • charter-context — Generate AI agent context files (CLAUDE.md, AGENTS.md, etc.)
  • project-charter — End-to-end orchestration of all three skills in sequence

📦 Installation

# CLI
uvx --from project-charter charter

# MCP Server
uvx --from "project-charter[mcp]" charter-mcp

# Full install via pip
pip install "project-charter[all]"

🔧 Requirements

  • Python >= 3.11
  • Works on Linux, macOS, and Windows

📄 Links

Built and maintained by Nextbridge.
If Project Charter helped your AI agent navigate your codebase safely, a ⭐ on GitHub would mean a lot!

Assets 2
Loading